sir, you have no idea how much you have helped me on SEM, without proper guidance it was extremely exhaustive but the ease with you explain this topic is truly appreciable again thanks a lot
I cannot stop watching your videos, they are more useful than the presential courses we are taking at the university. Thank you very much. I appreciate your help.
Can I just say you have literally saved my Master's degree by using your explanations and helpful videos. You have such a great way of explaining each part of data analysis. No other video I have watched has sunk in but watching yours have been course saving. Thank you so much for your support and help during this part of my studies.
I started listening to your lectures on AMOS from Lecture 1. I am not tired, but always glad listening to how simply you explain things. Engaging online listeners is very difficult, but you are exceptional. Indeed, you are a great prof, you know how to simplify things for people to learn. Thank you so much. I wish we meet one day in person to officially say DANKE (to wit, thank you). Now I can write my PhD thesis analysis. Greetings from Germany.
i started to listen your session, it was simply superb and the explanations were great. it helps me a lot in my PhD research work. i appreciate your good work and best wishes for upcoming days in different topics
This video was so helpful! I was lost as to what to do with this output for my dissertation as it has been a while since I was in AMOS. THANK YOU THANK YOU THANK YOU!
Hello sir, your videos are simply awesome. You are explaining very clearly, and it is easy to follow. Expecting more videos in future. Thank you sir for this effort.
Hello sir, I'm a new subscriber, and I found your explanations very comprehensive. Your videos helped me a lot in my research. I'm looking forward to learn more about research from your videos. Thank you!
Thanks. I am glad you liked it. Please find below the link to the Playlists SPSS: ruclips.net/p/PLb7vm6tsQ3KtPV0aJuJvLg6R2bcjvelCD AMOS: ruclips.net/p/PLb7vm6tsQ3Kvinj9MFF6wbQFDBdkU1ERP
Thanks for your comments. Here is the list of tutorials on Mediation Concept of Mediation ruclips.net/video/09Kmzy7kDNI/видео.html Simple Mediation ruclips.net/video/i44gFyqYDQI/видео.html Multiple Mediators ruclips.net/video/IVJTZi-Rd7I/видео.html Serial Mediation ruclips.net/video/ftX3dGLpn2o/видео.html
Hi, Fawad, thank you so much for your video. Recently I've watched some of your videos regarding structural equation modeling. It is so concise and easy to understand. Thank you! I have one question: what is the difference between this model fit in full structural model assessment and the model fit in confirmatory factor analysis? When I was reading some literature, there are some one-factor model, two-factor model, three-factor model in confirmatory factor analysis. And those researchers made comparison. I don't know what did they mean and why don't you include them? This was my first time getting to know CFA and SEM, sorry for my stupidity.
Pleasure. I am glad you liked it. Do share in your research circle. One Factor, Two, or Three, is assessing which factor model is best. However, experts do not recommend it, as factor structure should be based on theory not the data itself. The model fit between measurement model and structural model can be different, since both model types are different with measurement model based on covariance between constructs, and structural model based on impact of exogenous on endogenous.
Hi@@researchwithfawad would i have to report both the model fit for both the measurement model and structural model in my research report? I know that the model fit in the measurement model is to finesse the model by removing observed variables that are not strongly correlated to the latent variables. So do I report the model fit for the measurement model to justify which observed variables I removed and then rerun the model fit for the structural model and report the new model fit?
Hi Sir, first of all thank you very much for your video. I have a quick question what if the result is negative but significant. How do we interpret it?
Thanks. I am glad you liked it. Negative is the sign that shows direction of relationship. If it is negative and significant that means there is a significantly negative relationship between the two variables
@@researchwithfawad Thank you Sir, so if there is a significantly negative relationship between the two variables. The hypothesis will be accept or reject, Sir
Respected Sir, I have watched full vedio, plz guid whether we should use this modified CFA data file for structural model and also for meditation analysis. Or after modification of structural model this modified structural data file will be used for mediation.plz your response will be great solution for me like your previous responses.
I have a question that you didn't use the Composite Factor Scores to develop a Structural Model but you used a Measurement model to develop a Structural Model along with Observeable items. What is the difference and which model should we report in our results?
Hello Sir..Thank you so much for your great video explanation. I want to ask about the model fit in structural model analysis, can we improve the model fit same like in CFA analysis? Since my model fit scores are not really good, especially SRMR score. Once again, thank you so much for your great work.
Hi Sir! Thank you so much for a very informative and clear discussions. You explained it very well that a newbie like me was able to understand. I have few more questions though. Can you please help me understand Sir what is wrong with my data. I had somehow quite a good model fit with GFI=.856; IFI, TLI,CFI >.90; RMSEA=.049, SRMR=.0597 (or can I really say that this can be considered as a good model already?). However, my AVE are less than .5 (range is between .402 - .491). I did a lot of modifications already and even removed some constructs to get the above model fit values but my AVE values are not good. What does this mean? or what can I do to remedy this? Another question Sir (so sorry I hope you don't mind), when I did my CFA, I only included reflective latent variables and I was able to get the model fit values above. But when I did the path analysis, I added observed variables (my independent variables) and when I run again, the model fit values changed and lowered GFI to .70 and other indices to .8. If I will report for the Model Fit values, can I report the results during the CFA analysis without the observed variables or I should report the model fit values I generated in the path analysis? I'm so sorry if this is too long, but hoping for your guidance Sir. Thank you very much and more power to you and your channel!
Hi. I am glad you liked it the video. Model fit is about how well your data fits the model so you have fit indices. Reliability statistics like CR and Validity Statistics like AVE do not relate to fits. They are just based on the data itself. One cannot say that if one is good the other has to be better as well. Check the data like outliers, loadings, any respondent misconduct for improvement. As for path analysis, i recommend full structural model, path analysis will have lower model fit in comparison to full structural model.
Thank you so much for providing valuable information sir...i have a question if we have to test a relationship between A, B, C where A influence B and B inturn influence C, in this case C is our dependent variable obviously but who will be our independent variable? cox B act as an independent variable for C and A act as an independent variable for B...so can we call both A and B as independent variable or only B as an independent variable and A as a predictor.....
Can you please tell me about which Unique enteries are considered along with number of parameters to be estimated to calculate degree of freedom in FULL Structural Model? D.F. = 'p(p+1)/2' minus 'No. of parameter to be estimated'. For p= unique enteries and Parmaters= Paths+Variances+Covariances+Error Variance But question is that do we consider to calculate 'Paths' for both "Observeable items and Latent Items"? What are number of Variances and Error Variances? I am confused.
Thank you so much sir for your detailed videos on AMOS.Sir i am working on a Model with 5 exogenous construct,one mediating variable and one endogenous construct.My model fit indices of both measurement and structural model are same.Can you please tell me why is it so????
Hi Sir, thank you so much for this informative video. How can I interpret a full structural model that the 2 main constructs are covariate only? there is no direct relationship since in my study, I want to know the association of the 2 main constructs. Will I have to look at the estimates in covariance, as the two variables are not found in the regression weights?
There is a difference between Measurement and Structural Model. Structural model is to substantiate the proposed hypotheses, whereas measurement model is to assess the quality criteria including Factor Loadings, Reliability, and Validity.
@@researchwithfawad How can I interpret a full structural model that the 2 main constructs are covariate only? there is no direct relationship since in my study, I want to know the association of the 2 main constructs. Will I have to look at the estimates in covariance, as the two variables are not found in the regression weights?
Sir, i have few questions What if our CFA value (model fit, reliability and validity) not meet the desired criteria even after required method used to improve model fit indics as you teach us in your earlier tutorial of Amos)? I run CFA in amos, model fit indices meet the desired criteria however convergent and discriminant validity is not upto the mark? How can we save Amos figure? Thank you
Well this depends on the type of issue. For instance if there is an issue with discriminant validity, you might have to merge construct. If there is an issue with Convergent Validity, we have to see how serious it is. If there are too serious issues with Model fit, you might haveto recollect data.
Hellow sir Your videos are awesome 👍 I have a question that my model only have independent varibales to measure, the dependent variable do not have any measuring items so how can i check hypothese ? Please reply
Hello, I'm trying to run (calculate estimates) the model and it says:'no working data has been specified. Please select data file from the file menu and specify a data file.'
Hi sir, I am working on a sem model wherein I have developed an e-readiness scale and then I am studying it’s impact on service quality through the HEdPERF scale, but I am not able to analyse its results. Model fit is good, but regression weight and standardised regression weight are not significant. What to do? Please help me.
Please I have a question, I did my SEM analysis and I am getting a negative relationship between adaptation requirement and its impact (negative impact) on farm. The unstandardized estimate is (-0.184), S.E (0.034) C.R. (-5,521) P value(***). The squared multiple correlation is 0.133. The model fit is all good. Is my model wrong? if not, How can I interpret?
@@researchwithfawad Thank you very much for the prompt reply. I was expecting a positive relationship but now I understand better. You are a good tutor and the best.
Sir, what to do when regression weights were more than 2 or 3? Mostly it should be around 1 but i have a variable which shows more than 2 and 3 with a low CR value?
@@researchwithfawad should a double headed arrow be added to independent variables (even if they are actually factors of the latent variable and are treated as independent variables)?
Sir I am facing one problem. I have created structural model and my results are also good but two construct of mine have only two items so how can I justify that
Do we need to report the model fit in structural model. And is that a problem if the model fit indices values of sem model are same as that of the cfa model…. Kindly respond
@@researchwithfawad Thank you sir for your response!! Is there a good scientific paper i could reference to prove this? (the fact that accepting a hypothesis is based on the p-value and not the beta value)
![[TF2] The Russian Bear Trap](http://i.ytimg.com/vi/YFNSwz2_rOA/mqdefault.jpg)
sir, you have no idea how much you have helped me on SEM, without proper guidance it was extremely exhaustive but the ease with you explain this topic is truly appreciable again thanks a lot
Thanks for watching.
I cannot stop watching your videos, they are more useful than the presential courses we are taking at the university. Thank you very much. I appreciate your help.
Glad you like them!
Can I just say you have literally saved my Master's degree by using your explanations and helpful videos. You have such a great way of explaining each part of data analysis. No other video I have watched has sunk in but watching yours have been course saving. Thank you so much for your support and help during this part of my studies.
Thanks. I am glad the session was helpful.
I started listening to your lectures on AMOS from Lecture 1. I am not tired, but always glad listening to how simply you explain things. Engaging online listeners is very difficult, but you are exceptional. Indeed, you are a great prof, you know how to simplify things for people to learn. Thank you so much. I wish we meet one day in person to officially say DANKE (to wit, thank you). Now I can write my PhD thesis analysis. Greetings from Germany.
Thanks. I am really glad you liked the sessions. Keep watching and do share in your research circle.
Assalamu'alaikum Sir, thankk you for the videos, really helpful and hope us granted barakah
Thanks for watching
i started to listen your session, it was simply superb and the explanations were great. it helps me a lot in my PhD research work. i appreciate your good work and best wishes for upcoming days in different topics
You are most welcome
This video was so helpful! I was lost as to what to do with this output for my dissertation as it has been a while since I was in AMOS. THANK YOU THANK YOU THANK YOU!
Pleasure. I ma happy to know that the video was helpful. Keep Watching.
Very informative video. You make research much easy to understand. Thank you so much sir 👍🙏
Most welcome
Really a great lecture..... I hv learnt every step of SEM from you sir, thank u so much sir🙏
Thanks. I am glad you liked it.
You are doing great work. Gratitude.
Much appreciated
Thank you so so much! You saved my life and my master thesis. Much appreciated!!
Thanks. I am glad you liked it
Sir, you're the best fit for this analysis.... Thank you again... I really engoyed following your lessions... Magical work...
Thanks. I am glad to hear that. Thank you. Please do circulate in your research circle.
You are most welcome
@@researchwithfawad sure...
Very informative and comprehensive tutorial.. Thank you so much Sir
Thanks. I hope the sessions are helpful in learning the software
Now you have a new subscriber
Thank you. Do share it in your circle.
Hello sir, your videos are simply awesome. You are explaining very clearly, and it is easy to follow. Expecting more videos in future.
Thank you sir for this effort.
Keep watching
Hello sir,
I'm a new subscriber, and I found your explanations very comprehensive. Your videos helped me a lot in my research. I'm looking forward to learn more about research from your videos. Thank you!
Thanks and welcome. I am glad the videos are helpful. Stay tuned. Hopefully, i will keep sharing new and helpful material on research.
wawooo amazing sir
Thanks a lot
U made my day. Thanks a lot.
So nice of you
Thank you so much. Stay blessed! New sub.
Thanks for watching. Ameen
Nce explanation.
If you can give the list of all videos on SPSS and Amos, will be helpful very much. Please share.
Thanks. I am glad you liked it.
Please find below the link to the Playlists
SPSS: ruclips.net/p/PLb7vm6tsQ3KtPV0aJuJvLg6R2bcjvelCD
AMOS: ruclips.net/p/PLb7vm6tsQ3Kvinj9MFF6wbQFDBdkU1ERP
@@researchwithfawad thank you very much
Hello Sir,
Thank you for the session...
I have a request if you could kindly teach us how to run mediation analysis using AMOS....
Thanks for your comments.
Here is the list of tutorials on Mediation
Concept of Mediation
ruclips.net/video/09Kmzy7kDNI/видео.html
Simple Mediation
ruclips.net/video/i44gFyqYDQI/видео.html
Multiple Mediators
ruclips.net/video/IVJTZi-Rd7I/видео.html
Serial Mediation
ruclips.net/video/ftX3dGLpn2o/видео.html
This is an art you do prof. amazing. is it reliable to use Full Structural Model for the measurement model and structural model all at once?
Thanks. I am glad you liked it. In AMOS both shall be kept separate.
Hi, Fawad, thank you so much for your video. Recently I've watched some of your videos regarding structural equation modeling. It is so concise and easy to understand. Thank you! I have one question: what is the difference between this model fit in full structural model assessment and the model fit in confirmatory factor analysis? When I was reading some literature, there are some one-factor model, two-factor model, three-factor model in confirmatory factor analysis. And those researchers made comparison. I don't know what did they mean and why don't you include them? This was my first time getting to know CFA and SEM, sorry for my stupidity.
Pleasure. I am glad you liked it. Do share in your research circle.
One Factor, Two, or Three, is assessing which factor model is best. However, experts do not recommend it, as factor structure should be based on theory not the data itself. The model fit between measurement model and structural model can be different, since both model types are different with measurement model based on covariance between constructs, and structural model based on impact of exogenous on endogenous.
Hi@@researchwithfawad would i have to report both the model fit for both the measurement model and structural model in my research report? I know that the model fit in the measurement model is to finesse the model by removing observed variables that are not strongly correlated to the latent variables. So do I report the model fit for the measurement model to justify which observed variables I removed and then rerun the model fit for the structural model and report the new model fit?
Sir, what should we prefer for our thesis analysis-Path analysis or full structural model analysis? Thank u.
Please analyze with full structural model.
Sir, Can the standardised regression weight be above 1.0 and is it right to report that? Thank you!
Hi Sir, first of all thank you very much for your video. I have a quick question what if the result is negative but significant. How do we interpret it?
Thanks. I am glad you liked it. Negative is the sign that shows direction of relationship. If it is negative and significant that means there is a significantly negative relationship between the two variables
@@researchwithfawad Thank you Sir, so if there is a significantly negative relationship between the two variables. The hypothesis will be accept or reject, Sir
Your video is really very helpful! Do you have any articles available online for me to read? Thank you!
You can search
Khawaja Fawad Latif in Google Scholar
Respected Sir, plz guid me can we use modification indices to improve our structural model after running CFA.
Dear Scholar
Please watch this video
ruclips.net/video/3Y7dOXINV2E/видео.html
Respected Sir, I have watched full vedio, plz guid whether we should use this modified CFA data file for structural model and also for meditation analysis. Or after modification of structural model this modified structural data file will be used for mediation.plz your response will be great solution for me like your previous responses.
I have a question that you didn't use the Composite Factor Scores to develop a Structural Model but you used a Measurement model to develop a Structural Model along with Observeable items. What is the difference and which model should we report in our results?
Full structural model is better than simple path model based on composite scores.
Hello Sir..Thank you so much for your great video explanation. I want to ask about the model fit in structural model analysis, can we improve the model fit same like in CFA analysis? Since my model fit scores are not really good, especially SRMR score. Once again, thank you so much for your great work.
Thanks for watching. Yes, you can improve the model fit as in CFA.
Hi Sir! Thank you so much for a very informative and clear discussions. You explained it very well that a newbie like me was able to understand. I have few more questions though. Can you please help me understand Sir what is wrong with my data. I had somehow quite a good model fit with GFI=.856; IFI, TLI,CFI >.90; RMSEA=.049, SRMR=.0597 (or can I really say that this can be considered as a good model already?). However, my AVE are less than .5 (range is between .402 - .491). I did a lot of modifications already and even removed some constructs to get the above model fit values but my AVE values are not good. What does this mean? or what can I do to remedy this?
Another question Sir (so sorry I hope you don't mind), when I did my CFA, I only included reflective latent variables and I was able to get the model fit values above. But when I did the path analysis, I added observed variables (my independent variables) and when I run again, the model fit values changed and lowered GFI to .70 and other indices to .8. If I will report for the Model Fit values, can I report the results during the CFA analysis without the observed variables or I should report the model fit values I generated in the path analysis?
I'm so sorry if this is too long, but hoping for your guidance Sir. Thank you very much and more power to you and your channel!
Hi. I am glad you liked it the video.
Model fit is about how well your data fits the model so you have fit indices.
Reliability statistics like CR and Validity Statistics like AVE do not relate to fits. They are just based on the data itself. One cannot say that if one is good the other has to be better as well. Check the data like outliers, loadings, any respondent misconduct for improvement.
As for path analysis, i recommend full structural model, path analysis will have lower model fit in comparison to full structural model.
Please guide how to calculate Mean and Standard Deviation of every variable through AMOS?
You may use SPSS or Excel to do so
Thank you so much for providing valuable information sir...i have a question if we have to test a relationship between A, B, C where A influence B and B inturn influence C, in this case C is our dependent variable obviously but who will be our independent variable? cox B act as an independent variable for C and A act as an independent variable for B...so can we call both A and B as independent variable or only B as an independent variable and A as a predictor.....
Thanks for your interest. B is a mediator in this case and a is iv.
@@researchwithfawad sir what if we don't want to test mediation?
Then only check for individual relationships.
Can you please tell me about which Unique enteries are considered along with number of parameters to be estimated to calculate degree of freedom in FULL Structural Model?
D.F. = 'p(p+1)/2' minus 'No. of parameter to be estimated'.
For p= unique enteries and Parmaters= Paths+Variances+Covariances+Error Variance
But question is that do we consider to calculate 'Paths' for both "Observeable items and Latent Items"? What are number of Variances and Error Variances? I am confused.
I am sorry but i am not into pure statistics, my expertise are primarily in using tools.
Thank you so much sir for your detailed videos on AMOS.Sir i am working on a Model with 5 exogenous construct,one mediating variable and one endogenous construct.My model fit indices of both measurement and structural model are same.Can you please tell me why is it so????
I may not be able to identify the issues unless i see the model. You may email the screenshot for both the model on my email
kh.fawad83@gmail.com
Ok sir....thank you so much...I will do that
Hi Sir, thank you so much for this informative video. How can I interpret a full structural model that the 2 main constructs are covariate only? there is no direct relationship since in my study, I want to know the association of the 2 main constructs. Will I have to look at the estimates in covariance, as the two variables are not found in the regression weights?
Is it possible to use the created measurement model as the full structural model?
There is a difference between Measurement and Structural Model. Structural model is to substantiate the proposed hypotheses, whereas measurement model is to assess the quality criteria including Factor Loadings, Reliability, and Validity.
@@researchwithfawad How can I interpret a full structural model that the 2 main constructs are covariate only? there is no direct relationship since in my study, I want to know the association of the 2 main constructs. Will I have to look at the estimates in covariance, as the two variables are not found in the regression weights?
Sir, i have few questions
What if our CFA value (model fit, reliability and validity) not meet the desired criteria even after required method used to improve model fit indics as you teach us in your earlier tutorial of Amos)?
I run CFA in amos, model fit indices meet the desired criteria however convergent and discriminant validity is not upto the mark?
How can we save Amos figure?
Thank you
Well this depends on the type of issue. For instance if there is an issue with discriminant validity, you might have to merge construct.
If there is an issue with Convergent Validity, we have to see how serious it is.
If there are too serious issues with Model fit, you might haveto recollect data.
Hellow sir
Your videos are awesome 👍
I have a question that my model only have independent varibales to measure, the dependent variable do not have any measuring items so how can i check hypothese ?
Please reply
Thanks. You can not check any hypothesis if you don't not have DV measured.
Hello, I'm trying to run (calculate estimates) the model and it says:'no working data has been specified. Please select data file from the file menu and specify a data file.'
You will need to select the data file.
Hi sir,
I am working on a sem model wherein I have developed an e-readiness scale and then I am studying it’s impact on service quality through the HEdPERF scale, but I am not able to analyse its results. Model fit is good, but regression weight and standardised regression weight are not significant. What to do? Please help me.
Please watch to improve the model fit.
ruclips.net/video/TW4HvT6cqNY/видео.html
Please I have a question, I did my SEM analysis and I am getting a negative relationship between adaptation requirement and its impact (negative impact) on farm. The unstandardized estimate is (-0.184), S.E (0.034) C.R. (-5,521) P value(***). The squared multiple correlation is 0.133. The model fit is all good. Is my model wrong? if not, How can I interpret?
Negative means a negative relationship? Did you expect a negative relationship? If not you will need to justify it in your discussion.
@@researchwithfawad Thank you very much for the prompt reply. I was expecting a positive relationship but now I understand better. You are a good tutor and the best.
@@researchwithfawad Please one more thing, how do I report the t value (C.R) which has negative sign in my case, is it significant?
Sir, what to do when regression weights were more than 2 or 3? Mostly it should be around 1 but i have a variable which shows more than 2 and 3 with a low CR value?
Please refer to Standardized Regression weights.
Sir, do we need to put covariance in our independent variables even though in our hypothesis/conceptual model we do not correlate those?
Yes please. The Exogenous constructs shall be linked via double headed arrow or one should draw covariance between them .
@@researchwithfawad should a double headed arrow be added to independent variables (even if they are actually factors of the latent variable and are treated as independent variables)?
Assalamualaikum Sir, r square (squared multiple correlation), is there any specified limit for it?
Please watch
ruclips.net/video/jN2KFy35178/видео.html
Sir I am facing one problem. I have created structural model and my results are also good but two construct of mine have only two items so how can I justify that
Please refer to the source of the questionnaire
Do we need to report the model fit in structural model.
And is that a problem if the model fit indices values of sem model are same as that of the cfa model….
Kindly respond
Yes, fit indices shall be report for both CFA and Structural Model. They may be similar.
Sir,
A strange phenomena occurred to me, all Fit indices are OK acceptable but GFI is .751.
Please advise.
Thank you.
Dont worry about it. No need to report it. Report other indices.
@@researchwithfawad Thank you sir. It was a great help. And appreciate your efforts. The sessions are too good.
Sir, if the standardized beta coefficient value comes greater than 1 then how to interpret the hypothesis?
Please watch
ruclips.net/video/tK9u0EnJe_I/видео.html
Sir, question what if the results shows negative but significant result sir, will it still be accepted?
If you propose a positive relationship and 8t turns out to he negative, the hypothesis is not supported.
Thank you sir
Can i know how to make GOF index appear in the path diagram?
Please watch
ruclips.net/video/NtcFaZ0_r-I/видео.html
One question sir
What if CR Value is above 1.96 and P value
Acceptance and Rejection shall be based on the P value. Beta value is weight of impact. If p value is
@@researchwithfawad ok sir thank you so much for your quick response 🙏
@@researchwithfawad Thank you sir for your response!! Is there a good scientific paper i could reference to prove this? (the fact that accepting a hypothesis is based on the p-value and not the beta value)
How to decide which construct are dependent and independent ?
Please watch this session
ruclips.net/video/EE-Qh_LWA_g/видео.html
@@researchwithfawad do you mean that the dependent variable must be numerical in this case?
Dear Prof. What is the reference for this analysis?
Joel E Collier book on Structural Equation Modeling using AMOS.
@@researchwithfawad many thanks !!
Sir how to give model fit indices values in worksheet itself sir?
Sorry, If you want Model Fit in Excel, you can export if from AMOS or Copy it from AMOS Output.
@@researchwithfawad Thank you sir
Don't tell sorry sir
I asked various questions
You had reply immediately that's why I am telling thank you sir
May i have a writing template for writing result of amos?
I am sorry but those templates are not available for now. You have to use the video.
Sir plz provide us
ppT document
Please visit the website for lecture notes
www.researchwithfawad.com
@@researchwithfawad thanks alot sir
Sir can you please mail some books , pdf on structural equation modeling ?
Ptof. Zain ud din has made his book available in pdf. Visit his Research profile for download
www.researchgate.net/profile/Zainudin-Awang